Well there is something to be said for both potential biases in those counts. First of all, those being hospitalized are obviously among the most severe cases. And presumably people with the mildest cases will potentially not even be diagnosed, so the mortality rates would be skewed higher.

And secondly, that more people have died than have recovered doesn't say much other than that it takes longer to recover than to die. If you are going to die, it would appear that it would happen sooner rather than later. Whereas if you aren't going to die, you may take awhile before you've been officially cured.

That isn't to say that is absolutely what is happening here. But both scenarios are fairly likely, and therefore should mediate some of the bleakness of those numbers for the time being.